Here's the question to ask them: is it possible, particularly in the line play, for a player to make a positive play without accounting for it with a negative elsewhere? As an example: My point of view is that every time a player gets around another player and makes a positive play there must be an accounting made of how he managed to do that. I see two possibilities to make that accounting: 1. an offensive player (or players) made a mistake or were physically outmatched, thus allowing the defensive player to make the play. This is very straightforward. You give the defensive player a plus and you give the offensive player(s) a corresponding negative that cancels the plus out. You've reached zero and accounted for the play. 2. the offensive scheme broke down and nobody is actually in space to handle the defensive player. The fault is not clearly definable or attributable to any single player. You give the defensive player a plus and you give the offensive team a corresponding negative that indicates that a team error occurred. You've reached zero and accounted for the play. Under the first possibility you are accounting for the contribution that individual players make to the overall team's effort. Under the second you are accounting for schematic lapses that cause problems. Between the two you reach zero and the system balances. Even in situations where it doesn't look likely that that it's possible to make this type of accounting it usually is. If a defensive back makes a great play and bats a ball away at the last second to prevent a reception you figure out which of the QB and receiver were most responsible for the opportunity and you assign your negative. In fact you could assign a negative on the play to the offensive lineman if the throw was hurried because his man came free and pressured the QB. The kind of analysis that they are trying to do is really not about watching the individual players and figuring out why a play worked or not. It's really about watching the whole field and figuring out what happened. Given the fact that their numbers don't zero out the odds are very strong that what they are doing essentially is picking out a few elements in each play and over-weighting them without accounting for everything else that occurred in the process.
I agree they must take the "zero" into the equation. I also agree they need to account for a schematic error. The The call is to blitz 7 guys and there are 5 wide outs then one is bound to be open. If a play is made to that receiver then no defensive back is at fault, so the make goes against the Scheme/DC/Team how ever you want to handle it. Waht they are trying for really needs to be run as a balance sheet. Thier numbers need a lot of work to be seen as a statisticly significant way to evaluate play .
I don' work for the site so I'm really just explaining it from how I see it, don't take this an official answer. Yes,but I believe point 1 that you stated happens more often than this situation.An example of a positive play by a defender with no corresponding negative for the offense: lets say a WR runs a great route worthy of a +0.5 grade,his route causes him to get open makes the catch.CB gets a negative grade. Nows lets say it the next game,he runs the exact same route just as well but the opposite CB covers him prefectly and as a result he isn't open.The CB gets a +1 grade for his coverage,however the WR will still get his +0.5 because his route was just as good.It would be inconsistent to not give the WR the same grade for his route becuase he still made a good effort but the play developed differently from the previous time. However what is more likely that a player makes a negative play which doesn't get a corresponding positive grade for the other team.A commen example might be a lineman who fails to recognise the blitz,allowing a defender to come in untouched to sack the QB.The o-linemen who should of picked him up will get a negative grade where as the sacker will probably get a 0 grade.Just to point out 0 means average,you didn't do anything bad but perhaps you weren't challenged or do anything noteworthy.It's an easy sack and shouldn't get judged like a sack that invloves a player collapsing the pocket or beating a man. The site is for rating players not teams,so there is no team error system like you suggested in point 2.Things such as playcalling and coaching strategy aren't graded,just the players and how well they done given their job for each play. They look at what every player is doing on every play,grade them on how they done.If you need to weight or tweak grades because of the actual play result or opposition then you will have inconsistent grades.Like I said it's about the players not the team,the score tells you who is the better team,this site trys to tell you who are the better players.
That right there is the flaw in the ratings system. Both players can not make a positive play on the same play. If the reciver is not open he is not doing his job. In ability to sell the route, or make his break quick enough, caused him to be covered like a blanket. He is dead as an option for the QB, so did not get his job done on this play and deserves a negative. Same goes for Linemen. If the Center is supposed to push the NT back, but keeps him even, its a negative. If the NT is supposed to penetrate but is held, it should be a negative. If the cancel each other out in the line you need to know what thier job is on the given play to grade it properly. The site is a good attempt, but the logic is flawed in the scoring system used.
This would take a change in culture. The Steelers have been talking about smashmouth football and getting bloody since about 1971. It was the defense lead by Joe Greene and Andy Ham that started that kind of talk and the culture has remained ever since. The team most known for rough football before that was the Cleveland Browns, who had come up with it on offense as a method of exploiting Jim Brown's magnificent skills. They would say they just needed to get their uniforms a little bloody and dirty and Brown would handle the rest. The Steelers evolution to that concept came directly from having to stare down the Browns every year and they took it to the next level.
RE: Andy Ham You combined Andy Russell and Jack Ham there. No worries. I would add that the Steelers actually had a punishing defense prior to the arrival of Chuck Noll, Greene, Russell, and Ham. The Steelers were usually under .500 in the 1950s and '60s, but they beat the hell out of teams. Led by Ernie Stautner (to this day, still the only Steeler to have had his number retired) and Jack Butler, the Steelers would often times lose the game but physically hurt their opponents.
Yeah thanks for catching that. I was in the shower thinking about it afterwards and realized I'd done that. The Steelers were a hard hitting team from the mid 50's onwards but the Browns were the big dogs and really dominated them for a long time. The balance shifted at just about the time of the merger and the Browns never got back on top again. Basically the Steelers started winning the battle in the trenches and what had been a really poor matchup for them (9-31 at one point) became a favorable one.
But now the grades are being subjected to the quality of opponent,the player has done the exact same thing both times but will get a different grade(I'm talkling strictly about route running,the WR probably gets an additional positive for the catfch).BY doing that then grades will favour players matched up against weak opposition. That situation both players will probably get a 0 grade.DTs don't penerate every play and o-linemen don't push back linemen every play.Unless the o-linemen whifs a block or gets pushed backwards,he won't get a negative play.Likewise he'll get a positive grade only if he pushes a guy backward or makes a good block at the point of attack.
How are you determine quality of opponent? These types of stats are an attempt to see how well a player is doing their job. You cannot do that if you are creating a weighted system with individual players being give a weight that you are pulling from thin air. Doing this throws in an extra uncontrolled variable. As an example, Say I believe that Revis is the best cornerback in the NFL, so I give him the maximum weight. Then he goes out and gets torched for 3 big plays in one game. This would bring into question my previous grade handed to him and possibly give too much weight to the guy who made the plays against him. That makes this grading scale not subjective an worthless as a means for judging player performance. Players should start at 0. Then impact on any play gives them a plus or minus from there. That way an Inferior WR who is covered all day would tally negatives on every play. The corresponding CB would receive positives for every play that WR. is shut down. Then you have a grade sheet that does not take into account what one person thinks the weight of a player should be, but the player receives a performance grade and the weight of the player's effectiveness is proven over time. You need to remove the implied human grading for it to have any weight in a debate.
